agentleFS
davidmatousekdavidmatousek/tachi/.claude/skills/~aod-build/SKILL.md

~aod-build

Skill95 starsChanged 38 days agoApache-2.0View on GitHub

Generate standardized checkpoint reports for multi-phase implementation projects. Use this skill when pausing implementation at strategic milestones (phase completion, user story completion, critical features) to create comprehensive progress reports with task breakdowns, metrics, knowledge base entries, and resume instructions.

The file

We split this skill into the rules it gives an agent

9 rules. About 0 are common, 8 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈8

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Create a checkpoint at:

    line 14
  2. Distinctive:

    Complete phase finish: - All tasks in a phase marked complete

    line 16
  3. Distinctive:

    Every checkpoint report includes:

    line 94
  4. Distinctive:

    Task range: Start and end task IDs

    line 111
  5. Distinctive:

    When prompted for knowledge base entries, provide learnings that:

    line 118
  6. Very specific:

    Update CHECKPOINTS_README.md with new checkpoint.

    line 206
  7. Distinctive:

    Create checkpoints proactively: - Don't wait until pausing

    line 306
  8. Distinctive:

    Capture knowledge fresh: - Document learnings immediately

    line 307
  9. Distinctive:

    Test resume commands: - Verify they work before distributing

    line 310

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Implementation Checkpoint Skill

Purpose

Automatically generate standardized checkpoint reports that document implementation progress at strategic milestones. Each report includes completed tasks, remaining work, technical achievements, knowledge base entries, quality metrics, and resume instructions.

When to Use

Create a checkpoint at:

  1. Complete phase finish - All tasks in a phase marked complete
  2. User story completion - Entire user story implemented and tested
  3. Critical milestone - Core functionality ready (e.g., MVP core)
  4. Before pausing - Stopping for review, testing, or user feedback
  5. After architect review - Major approval/sign-off received

How to Use

Quick Usage

# Navigate to skill directory
cd .claude/skills/~aod-build

# Generate checkpoint (interactive prompts will follow)
python scripts/generate_checkpoint.py \
  --tasks-file ../../specs/{feature-id}/tasks.md \
  --output-dir ../../specs/{feature-id} \
  --description "{Phase-Description}"

Step-by-Step Process

  1. Analyze completion status
    # Script automatically:
    # - Reads tasks.md
    # - Counts [X] completed vs [ ] pending
    # - Calculates progress percentage
    # - Identifies completed phases
    # - Determines task range
  2. Provide context when prompted
    Key achievements (3-5 bullet points):
    > - Achievement 1
    > - Achievement 2
    > - Achievement 3
    
    Knowledge base entries captured (count or 'none'):
    > 2
    
    [For each KB entry, provide: Title, Problem, Solution]
  3. Review generated report
    # Output: CHECKPOINT-{NN}_{Description}_{TaskRange}.md
    # Location: specs/{feature-id}/
  4. Verify index updated
    # Check: CHECKPOINTS_README.md includes new checkpoint

Checkpoint Naming Convention

Format: CHECKPOINT-{NN}_{Phase-Description}_{TaskRange}.md

Components:

  • {NN}: Sequential checkpoint number (01, 02, 03...) - auto-detected
  • {Phase-Description}: Milestone name (e.g., "Phase1-2-Complete", "US1-Complete")
  • {TaskRange}: Task IDs covered (e.g., "T001-T030")

Examples:

✓ CHECKPOINT-01_Phase1-2-Complete_T001-T030.md
✓ CHECKPOINT-02_US1-Complete_T001-T039.md
✓ CHECKPOINT-03_MVP-Ready_T001-T101.md
✓ CHECKPOINT-04_ProductionReady_T001-T120.md

See references/naming_conventions.md for detailed rules.

Required Sections

Every checkpoint report includes:

  1. Executive Summary - What was completed, what's next
  2. Completion Status - Progress table with percentages
  3. What Was Implemented - Detailed task breakdown by phase
  4. Key Technical Achievements - Highlights and innovations
  5. Files Created/Modified - Complete file inventory with counts
  6. Knowledge Base Captured - Critical learnings documented
  7. Server/Database Status - Current state of running services
  8. Quality Metrics - Code quality, reviews, test coverage
  9. Remaining Work - What's left and estimated time
  10. Resume Instructions - How to continue implementation

Metrics Calculated Automatically

  • Progress percentage: (completed / total) * 100
  • Phase completion: Per-phase progress tracking
  • Task range: Start and end task IDs
  • Next task: First pending task to work on

See references/metrics_formulas.md for formulas and examples.

Knowledge Base Integration

When prompted for knowledge base entries, provide learnings that:

✅ DO include:

  • Solved problems that took >30 minutes
  • Non-obvious gotchas or edge cases
  • Architectural decisions with trade-offs
  • Patterns that should be reused
  • Issues that could happen again

❌ DON'T include:

  • Obvious facts ("we use PostgreSQL")
  • Generic statements ("testing is important")
  • Implementation progress ("completed task T026")
  • Temporary workarounds

Output Files

The skill generates/updates:

  1. Checkpoint Report: CHECKPOINT-{NN}_{Description}_{TaskRange}.md
    • Location: specs/{feature-id}/
    • Size: ~15-20 KB
    • Format: Markdown with tables and code blocks
  2. Checkpoints Index: CHECKPOINTS_README.md (updated)
    • Location: specs/{feature-id}/
    • Adds new checkpoint entry
    • Updates progress tracking table

Resume Commands Generated

The checkpoint report includes auto-generated resume commands:

# For team-lead orchestration
/aod.build --resume

# For direct continuation
# Continue from task {NEXT_TASK}

Scripts Reference

analyze_tasks.py

Parse tasks.md and extract metrics.

Input: Path to tasks.md file Output: JSON with completion metrics

Usage:

python scripts/analyze_tasks.py specs/001-{{PROJECT_NAME}}/tasks.md

Output Format:

{
  "total_tasks": 120,
  "completed_tasks": 32,
  "progress_percentage": 26.67,
  "phases": [...],
  "task_range": "T001-T030",
  "next_task": "T031"
}

generate_checkpoint.py

Generate checkpoint report from template.

Usage:

python scripts/generate_checkpoint.py \
  --tasks-file specs/{feature-id}/tasks.md \
  --output-dir specs/{feature-id} \
  --description "{Phase-Description}"

Interactive Prompts:

  • Key achievements summary
  • Knowledge base entries (optional)
  • Per-entry: title, problem, solution

Output: Complete checkpoint report file

update_index.py

Update CHECKPOINTS_README.md with new checkpoint.

Usage:

python scripts/update_index.py \
  --checkpoints-readme specs/{feature-id}/CHECKPOINTS_README.md \
  --checkpoint-file CHECKPOINT-{NN}_{Description}_{TaskRange}.md \
  --checkpoint-num {NN}

Changes Made:

  • Adds checkpoint to "Active Checkpoints" section
  • Updates progress tracking table
  • Updates "Next Checkpoint Planned" footer

Templates Reference

checkpoint_template.md

Complete Markdown template for checkpoint reports. Contains all required sections with template variables for substitution.

Location: references/checkpoint_template.md

Key Variables:

  • {CHECKPOINT_NUM}, {PHASE_DESCRIPTION}, {TASK_RANGE}
  • {DATE}, {FEATURE_ID}, {TOTAL_TASKS}, {COMPLETED_TASKS}
  • {PROGRESS_PERCENTAGE}, {PHASES_TABLE}, {KEY_ACHIEVEMENTS}
  • {KB_ENTRIES}, {NEXT_TASK}, {RESUME_COMMAND}

CHECKPOINTS_README_template.md

Template for creating the checkpoints index file (first-time use).

Location: assets/CHECKPOINTS_README_template.md

Usage: Copy to specs/{feature-id}/CHECKPOINTS_README.md when creating first checkpoint.

Examples

Example 1: After Completing Phase 1-2

$ python scripts/generate_checkpoint.py \
    --tasks-file specs/001-{{PROJECT_NAME}}/tasks.md \
    --output-dir specs/001-{{PROJECT_NAME}} \
    --description "Phase1-2-Complete"

Analyzing tasks.md...
✓ Found 120 total tasks
✓ Found 32 completed tasks (27%)
✓ Task range: T001-T030
✓ Checkpoint #01 (auto-detected)

Key achievements (3-5 bullet points):
> - Database schema with 6 entities, 19 indexes
> - Architect review APPROVED
> - MVP core: Atomic locking with row-level locks

Knowledge base entries: 5

[... interactive prompts for KB entries ...]

✓ Generated: CHECKPOINT-01_Phase1-2-Complete_T001-T030.md
✓ Updated: CHECKPOINTS_README.md

Next: Resume with /aod.build --resume

Example 2: After Completing User Story 1

$ python scripts/generate_checkpoint.py \
    --tasks-file specs/001-{{PROJECT_NAME}}/tasks.md \
    --output-dir specs/001-{{PROJECT_NAME}} \
    --description "US1-Complete"

✓ Checkpoint #02: US1-Complete (T001-T039)
✓ Progress: 33% (39/120 tasks)

[... prompts and generation ...]

✓ Generated: CHECKPOINT-02_US1-Complete_T001-T039.md

Troubleshooting

Q: Script can't find tasks.md A: Provide absolute path or verify relative path from skill directory

Q: Checkpoint number is wrong A: Manually specify with --checkpoint-num NN flag

Q: Template variables not substituted A: Check template file exists in references/checkpoint_template.md

Q: Index file not updating A: Verify CHECKPOINTS_README.md exists in output directory

Best Practices

  1. Create checkpoints proactively - Don't wait until pausing
  2. Capture knowledge fresh - Document learnings immediately
  3. Be specific in achievements - "Implemented X" not "Made progress"
  4. Validate metrics - Review auto-calculated percentages
  5. Test resume commands - Verify they work before distributing

Skill Version: 1.0.0 Last Updated: 2025-11-20 Maintainer: Team Lead Agent


Common Rationalizations

RationalizationReality
"I'll skip the checkpoint — I'm only mid-phase"Step 1 (line 17) lists critical milestones too; mid-phase pause without report leaves no resume artifact.
"KB entries can wait until after delivery"Line 307 directs capturing knowledge fresh; deferred entries miss the gotchas described in lines 121-126.
"Achievements like 'made progress' are fine"Line 308 requires specific achievements; vague entries violate the KB DON'T list (line 130).
"I can hand-write achievements without the script"Quick Usage (line 27) requires generate_checkpoint.py — manual reports skip the auto progress percentage from line 109.
"Resume commands are documentation, not testing"Line 311 requires testing resume commands before distributing; untested commands break next session.

Red Flags

  • Agent creates a checkpoint without running analyze_tasks.py per the metrics block (lines 162-183).
  • Agent's KB entries match the DON'T list (line 128): "obvious facts," "generic statements," "implementation progress."
  • Agent skips line 310 (validate metrics): publishes auto-calculated percentages without sanity-checking them.
  • Agent omits the Required Sections list (line 96) — Quality Metrics, Resume Instructions, or Files Created.
  • Agent uses arbitrary checkpoint numbering instead of the auto-detected {NN} from the script (line 78).
  • Agent's report contains template variables like {NEXT_TASK} unsubstituted — Troubleshooting Q3 (line 298) flags this.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.